DESCRIPTION/ABSTRACT: Stroke recurrence is a major public health problem. It has been understudied epidemiologically despite comprising ~25% of all stroke in the US, and leading to higher morbidity and mortality compared to initial events. Furthermore, compared to our understanding of the role of traditional risk factors for first stroke, we have limited understanding of the risk factors for recurrent stroke. Radiological measures of small vessel and large-vessel disease may provide a measure of a subject?s current ?Brain Health? to predict future cerebrovascular events and vascular dementia. Understanding small vessel disease in particular?manifesting as white matter hyperintensity, microbleeds, lacunar infarcts, dilated perivascular spaces, and brain atrophy?is a top NINDS priority. Although the Framingham risk score is widely used for clinical decision-making, and the ABCD scores predict 90-day stroke risk among minor stroke/TIA patients, no clinical prediction tool in a broader ischemic or a hemorrhagic stroke population-based cohort has been established. A Framingham-like ?Risk Function Tool for Secondary Stroke,? especially one incorporating the full gamut of modern imaging parameters developed in a racially diverse population, would likely identify high-risk patients for targeted education interventions, novel antithrombotics, and aggressive clinical care. Moreover, it may reveal novel causal pathways and risk factors for future investigation. We propose an ancillary study to the ongoing Greater Cincinnati/Northern Kentucky Stroke Study (GCKNSS) entitled APRISE (Assessing Population-based Radiological brain health in Stroke Epidemiology). Our objective is to build on the GCNKSS infrastructure to radiologically characterize the full stroke/TIA population of Greater Cincinnati/Northern Kentucky from 2015 and create the first modern, largely MRI-based characterization of brain health in a contemporary stroke/TIA population. Our specific aim is to create a prediction model of cerebrovascular disease recurrence (ischemic or hemorrhagic stroke) incorporating imaging parameters in a biracial, large-scale ischemic stroke/TIA population using state-of-the-art modeling approaches. We will achieve these goals by radiologically characterizing the stroke/TIA population (~2700 ischemic and ~525 hemorrhagic stroke, and ~1025 TIA patients) in the NIH-funded GCNKSS of a 1.4 million, biracial population. We will leverage the extensive GCNKSS infrastructure, which is already in place to detect recurrent stroke, placing us in a unique position to develop a recurrent stroke risk model using newly acquired radiologic data with remarkable efficiency. The proposed study will allow the first and only population-based characterization of the entire imaging spectrum of brain health in stroke/TIA patients. Our multi-disciplinary team with imaging, epidemiology, outcomes, and biostatistics expertise will address NINDS priority research areas with precision and innovation currently unavailable.